Prepare the input
Paste finished prose rather than isolated punctuation. Keep abbreviations, quotes, and line breaks as they will appear, because all can affect segmentation.
Inspect sentence structure locally
Estimate sentence boundaries and report words and characters alongside the total, using browser language segmentation where available and no text upload.
Counting and transformations run locally. Nirmion does not send the text in this workspace to an API.
Product guide
Paste finished prose rather than isolated punctuation. Keep abbreviations, quotes, and line breaks as they will appear, because all can affect segmentation.
There is no hidden grammar score. The tool estimates sentence units and reports word and character context. It does not label a sentence as correct, complete, active, passive, simple, or complex.
Use the total to compare sections, calculate an average manually, or spot an unexpectedly dense block. Review dialogue, bullet lists, initials, decimal numbers, and abbreviations where automated boundaries can differ from editorial intent.
How it works
Supported browsers segment the source with sentence granularity and retain only segments containing a Unicode letter or number. The fallback separates on common Western and full-width sentence punctuation or line boundaries. That provides a useful deterministic estimate without claiming a grammar parse or language model analysis.
Useful when
Sentence totals are useful when paired with the actual prose and a clear editorial goal.
Questions answered
Practical boundaries matter with text tools. These answers explain what is measured, what is retained, and when a manual review is still needed.
Yes when it forms a detected segment containing letters or numbers. Exclamations and common full-width punctuation are handled similarly.
Boundary detection uses language rules available in the browser, but abbreviations and domain-specific notation remain ambiguous. A fallback based on punctuation is less context-aware, so manual review is necessary.
No. It measures likely sentence boundaries only. It does not diagnose fragments, run-ons, punctuation errors, or writing quality.
Privacy and retention
Nirmion performs these operations with browser JavaScript. The workspace does not call the catalogue API with your source text, and completed-job content is kept only in this component memory for re-download.
A downloaded file remains in your browser downloads, device storage, backups, and any location where you share it. Clear or close the page when working on a shared device, and avoid pasting secrets into extensions or third-party clipboard managers.
Tool details
Nirmion Sentence Counter evaluates up to 500,000 characters locally and reports likely sentences together with words and grapheme characters. This makes the result useful for structural comparison, readability planning, transcript cleanup, and editorial review without presenting the count as a grammar score.
The downloadable report is text_nirmion_tools_sentence_report.txt. Browser segmentation improves multilingual boundary handling where supported, but abbreviations and unconventional formats can remain ambiguous, so the page documents those limitations clearly.